In this article, we aim at developing neighborhood-based neural models for link prediction. We design a novel multispace neighbor attention mechanism to extract universal neighborhood features by capturing latent importance of neighbors and selectively aggregate their features in multiple latent spaces. Grounded on this mechanism, we propose two link prediction models, i.e., self neighborhood attention network (SNAN), which predicts the link of two nodes by encoding and matching their respective neighborhood information, and its extension cross neighborhood attention network (CNAN), where we additionally design a cross neighborhood attention to directly capture structural interactions between two nodes. Another key novelty of this work is that we propose an adversarial learning framework, where a negative sample generator is devised to improve the optimization of the proposed link prediction models by continuously providing highly informative negative samples in the adversarial game. We evaluate our models with extensive experiments on 12 benchmark data sets against 14 popular and state-of-the-art link prediction approaches. The results strongly demonstrate the significant and universal superiority of our models on various types of networks. #link# The effectiveness and robustness of the proposed attention mechanism and adversarial learning framework are also verified by detailed ablation studies.The rapid development of deep learning algorithms provides us an opportunity to better understand the complexity in engineering systems, such as the smart grid. Most of the existing data-driven predictive models are trained using historical data and fixed during the execution stage, which cannot adapt well to real-time data. In this research, we propose a novel online meta-learning (OML) algorithm to continuously adapt pretrained base-learner through efficiently digesting real-time data to adaptively control the base-learner parameters using meta-optimizer. The simulation results show that 1) both ML and OML can perform significantly better than online base learning. 2) OML can perform better than ML and online base learning when the training data are limited, or the training and real-time data have very different time-variant patterns.This work focuses on robust speech recognition in air traffic control (ATC) by designing a novel processing paradigm to integrate multilingual speech recognition into a single framework using three cascaded modules an acoustic model (AM), a pronunciation model (PM), and a language model (LM). The AM converts ATC speech into phoneme-based text sequences that the PM then translates into a word-based sequence, which is the ultimate goal of this research. The LM corrects both phoneme- and word-based errors in the decoding results. The AM, including the convolutional neural network (CNN) and recurrent neural network (RNN), considers the spatial and temporal dependences of the speech features and is trained by the connectionist temporal classification loss. To cope with radio transmission noise and diversity among speakers, a multiscale CNN architecture is proposed to fit the diverse data distributions and improve the performance. Phoneme-to-word translation is addressed via a proposed machine translation PM with an encoder-decoder architecture. RNN-based LMs are trained to consider the code-switching specificity of the ATC speech by building dependences with common words. We validate the proposed approach using large amounts of real Chinese and English ATC recordings and achieve a 3.95% label error rate on Chinese characters and English words, outperforming other popular approaches. The decoding efficiency is also comparable to that of the end-to-end model, and its generalizability is validated on several open corpora, making it suitable for real-time approaches to further support ATC applications, such as ATC prediction and safety checking.Understanding how neural networks learn remains one of the central challenges in machine learning research. From random at the start of training, the weights of a neural network evolve in such a way as to be able to perform a variety of tasks, such as classifying images. Here, we study the emergence of structure in the weights by applying methods from topological data analysis. We train simple feedforward neural networks on the MNIST data set and monitor the evolution of the weights. When initialized to zero, the weights follow trajectories that branch off recurrently, thus generating trees that describe the growth of the effective capacity of each layer. When initialized to tiny random values, the weights evolve smoothly along 2-D surfaces. We show that natural coordinates on these learning surfaces correspond to important factors of variation.In this article, a model-free online adaptive dynamic programming (ADP) approach is developed for solving the optimal control problem of nonaffine nonlinear systems. Combining the off-policy learning mechanism with the parallel paradigm, multithread agents are employed to collect the transitions by interacting with the environment that significantly augments the number of sampled data. On the other hand, each thread agent explores the environment with different initial states under its own behavior policy that enhances the exploration capability and alleviates the correlation between the sampled data. After the policy evaluation process, only one step update is required for policy improvement based on the policy gradient method. The stability of the system under iterative control laws is guaranteed. Moreover, the convergence analysis is given to prove that the iterative Q-function is monotonically nonincreasing and finally converges to the solution of the Hamilton-Jacobi-Bellman (HJB) equation. For implementing the algorithm, the actor-critic (AC) structure is utilized with two neural networks (NNs) to approximate the Q-function and the control policy. Finally, the effectiveness of the proposed algorithm is verified by two numerical examples.The common spatial pattern (CSP) algorithm is a well-recognized spatial filtering method for feature extraction in motor imagery (MI)-based brain-computer interfaces (BCIs). However, due to the influence of nonstationary in electroencephalography (EEG) and inherent defects of the CSP objective function, the spatial filters, and their corresponding features are not necessarily optimal in the feature space used within CSP. In this work, we design a new feature selection method to address this issue by selecting features based on an improved objective function. Especially, improvements are made in suppressing outliers and discovering features with larger interclass distances. Moreover, https://www.selleckchem.com/products/arv-771.html based on the Dempster-Shafer theory is proposed, which takes into consideration the distribution of features. With two competition data sets, we first evaluate the performance of the improved objective functions in terms of classification accuracy, feature distribution, and embeddability. Then, a comparison with other feature selection methods is carried out in both accuracy and computational time.